.
Angewandte
Communications
much faster even at 08C in iPr2NH (Figure 1b, entries 3 and
4). However, Et3N was less effective under the same
conditions (Figure 1b, entry 5). It was noteworthy that these
three Pd0-catalyzed reactions all exhibited zero-order kinet-
ics, which apparently indicated that reductive elimination
might be the rate-limiting step in this transformation.
Detailed kinetic investigations of the Sonogashira cou-
pling reactions were further performed in iPr2NH at 08C
[Eq. (1) in Figure 2]. In Figure 2a the kinetic profiles at
varying initial concentrations of 1a are shown. The nearly
identical initial rates indicated that this reaction was zero-
order dependent on c(1a). This result could rule out the
oxidative addition of ArI with the Pd catalyst as the rate-
limiting step. Meanwhile, Figure 2b showed the kinetic plots
of the reactions when c(2a) was varied. The almost perfectly
overlaying initial kinetic rates revealed that this Sonogashira
coupling was also independent on the concentration of the
terminal alkyne c(2a). Furthermore, kinetic measurements
with different concentrations of the [Pd] catalyst were
investigated. A linear relationship was observed in Figure 2c
when plotting the initial rate versus c([Pd]). This result
suggested that the reaction was first-order dependent on the
concentration of the Pd catalyst in the tested Pd concen-
tration range. Thus far, all the above kinetic observations
suggested that reductive elimination might be the rate-
limiting step.
However, when we tried to further reveal the relationship
between the electronic effect of substituent and rate-limiting
step, we faced a puzzle. As shown in Figure 3a, the more
electron-withdrawing the substituent on the electrophile ArI
1 was, the faster the reaction proceeded. This result was
unanticipated. The reversed trend was observed when the
substituents on the nucleophile alkynes 2 were investigated
(Figure 4a). The more electron-withdrawing group the alkyne
had, the slower the reaction occurred. Although the elec-
Figure 3. a) Kinetic profiles of different aryl iodides 1 monitored by
in situ IR spectroscopy. b) The Hammett correlation plot. The rate
constant k was obtained by fitting the linear plot of c(3) versus time.
Plotting log(k/k0) versus the substituent constant sp for these sub-
strates gave a linear relation (correlation coefficient r2 =0.96, positive
slope 1=1.29).
bond. At the beginning, we assumed if the reductive
elimination is rate-limiting, the electronic effects on varying
the substituents of only one, the electrophile or the nucleo-
phile, should follow the same trend when the other is
constant. The results of the experiments shown in Figures 3a
and 4a were just different from the speculation. This puzzle
enforced us to reappraise the catalytic cycle of the Sonoga-
shira coupling reaction.
In fact, different from other Pd-catalyzed cross-coupling
reactions, the Sonogashira coupling usually also required Cu
salt as a cocatalyst.[7] In the above-mentioned kinetic experi-
ments, CuI (2 mol%) was also employed. Two synergistic
catalytic cycles are generally proposed for the Sonogashira
coupling reaction (Scheme 1).[1h,j,l,m] For the Pd-catalyzed
Sonogashira coupling, one usually considers that ArI is the
electrophile, and the terminal alkyne is the nucleophile. In
this way, when the kinetic data showed the reaction was
independent on both c(electrophile ArI) and c(nucleophile
alkyne), reductive elimination
À
tronic effects on the reductive elimination to form a C C
bond had been revealed to some extent,[6] there are rare
À
examples for the reductive elimination to form a Csp Csp
2
as the rate-limiting step could
be easily envisioned. However,
a careful analysis of the cata-
lytic cycle reveals the terminal
alkyne is not the direct nucleo-
phile of the Pd-catalyzed cycle!
The proposed copper–acetylide
is the nucleophile for the Pd-
catalyzed cycle, and is also the
catalytic species for the Cu-
catalyzed cycle.[8] Therefore,
the zero-order kinetic depend-
ences of c(ArI) and c(alkyne)
cannot suggest reductive elim-
ination is the rate-limiting step
Figure 2. a) Kinetic profiles of different initial concentrations of 1a (from 0.2m to 0.4m). Reaction
conditions: 2a (0.22m), [Pd(PPh3)4] (0.002m), CuI (0.004m), carried out in iPr2NH (5 mL) at 08C.
b) Kinetic profiles of different initial concentrations of 2a (from 0.22m to 0.4m). Reaction conditions: 1a
(0.20m), [Pd(PPh3)4] (0.002m), CuI (0.004m), carried out in iPr2NH (5 mL) at 08C. c) Kinetic profiles of
different [Pd(PPh3)4] loadings (from 0.25 mol% to 1.0 mol%) with CuI (0.004m), 1a (0.20m), and 2a
(0.22m), carried out in iPr2NH (5 mL) at 08C. Reactions were monitored by in situ IR spectroscopy.
in this case. The rate-limiting
transmetalation of copper–ace-
tylide with the R1PdX complex
A could also lead to the zero-
order kinetic behavior.
2
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2012, 51, 1 – 5
These are not the final page numbers!